Typical Scenarios Selection Of Wind Power Output Based On Clustering Algorithm

The output of wind power station varies with wind energy,the influence of weather,season,region make the outputs have the characteristics of strong randomness,volatility and intermittent.Because the increasing capacity of wind farm will influence real-time balancing of power,severe challenges are proposed,the power system requires the integration capability of wind power when planning and operation.It is necessary to select a typical wind power scene for the affordable consumption level of wind power and the stability of the power grid.According to the wind power scene,a typical output characteristic of wind power in the region could be achieved for the design of regional power grid,assessment of wind power acceptance,optimal allocation of reactive power and operation and scheduling planning of regional grid with wind power.In order to describe the characteristics of wind power output in the region,an improved fuzzy C-means clustering algorithm is proposed for typical scene selection of wind power in the region.Firstly,initialize the wind power sample clustering center and the preset scene partition number.Based on this,the traditional fuzzy C means clustering algorithm is applied to cluster analysis of the wind power output samples,and the preliminary clustering results are obtained.Then considering the division quality of typical scenarios of wind power,used the validity function XB which considers both the compactness and interclass separation as the basis of determining the optimal number of clusters,so as to achieve the best partition of wind power output samples.According to the clustering results,it could extract the typical scene of regional wind power with the largest probability of occurrence and the typical scene of regional wind power with the most volatility.Aiming at the uncertaninty of the fuzzy C-mean clustering algorithm which it is necessary to set the number of scenes in advance.A hierarchical clustering algorithm is proposed to select typical scenarios of wind power in regional power grid.Then,in order to improve the quality of wind power typical scenarios' selection,the sum of squares of deviations is used to describe the difference between interclass samples,which is regarded as a basis to determine the number of clusters,and it realizes the samples'effective division.According to the clustering results,it could extract the typical scene of regional wind power with the largest probability of occurrence and the typical scene of regional wind power with the most volatility.This paper uses actual operation data of wind power in a region to carry out the simulation of two selecting method for the typical scene of the regional wind power mentioned above,which verifies the feasibility of the method.The achievement of this paper could serve the theoretical basis for the selection of typical scene of the regional wind power.